Does money earned abroad build anything lasting at home?
Abundance from Abroad: Migrant Income and Long-Run Economic Development
Gaurav Khanna et al. (2026) · Review of Economic Studies
Replication data ↗ (openICPSR)
The archive requires an account we do not hold, so the files were not opened. What follows rests on the published abstract and on this platform's own observations.
What the authors found
Khanna, Murathanoglu, Theoharides and Yang use the 1997 Asian Financial Crisis, which moved exchange rates differently for Philippine provinces depending on where their migrants worked, as a shift-share instrument. Initial migrant income shocks were magnified roughly sixfold over the following decades through higher domestic income, more schooling, and more skilled migration. Strikingly, 73.6 per cent of the long-run income gain came from domestic rather than migrant income, and about a fifth of it operated through educational investment.
What our data adds
The Philippines remains among the most remittance-dependent large economies in this database, receiving 41.6 billion dollars a year, worth 8.5 per cent of GDP — over five times its foreign direct investment. If the multiplier the authors identify is anywhere near right, that flow is not merely 8.5 per cent of current income but a claim on future domestic output through the schooling it finances. The most remittance-dependent economy of all is Tajikistan at 57.7 per cent of GDP, followed by Honduras at 30.1 and El Salvador at 27.5.
The Philippines against the most remittance-dependent economies.
| Figure | Value | Indicator |
|---|---|---|
| Philippines, remittances received | 41.6bn USD | BX.TRF.PWKR.CD.DT |
| As a share of GDP | 8.5% | BX.TRF.PWKR.DT.GD.ZS |
| Tajikistan, the most dependent economy | 57.7% | BX.TRF.PWKR.DT.GD.ZS |
| Share of long-run gain from domestic income (paper) | 73.6% | BX.TRF.PWKR.CD.DT |
Every figure here is computed from observations in this database and can be charted, downloaded or queried.
What this cannot show
The sixfold magnification is identified from variation across Philippine provinces around one exchange-rate event, and there is no basis for assuming it transfers to Tajikistan or Honduras, whose migration corridors, education systems and labour markets differ in every relevant respect. Applying the multiplier to a national ratio would be arithmetic dressed as inference. The national series cannot see schooling, skill composition or the domestic-versus-migrant split that is the paper's headline result.
The data behind it
Personal remittances, received (current US$)
BX.TRF.PWKR.CD.DT
WBPersonal remittances, received (% of GDP)
BX.TRF.PWKR.DT.GD.ZS